Ankle fracture and ligament injury diagnosis method based on machine learning

Abstract

The invention discloses an ankle fracture and ligament injury diagnosis method based on machine learning. The method comprises the steps of determining a computer feature extraction region, marking the classification of all X-ray image fracture types in each region, and marking the ligament injury type of a patient corresponding to an X-ray image, i.e., injury and non-injury, preprocessing the X-ray image to obtain a foreground image of the ankle skeleton, building an ankle fracture prediction model based on a BoVW+SVM algorithm, searching a frequent item set between bone injuries and ligamentinjuries in a case through an Apriori algorithm, mining association rules between affairs, taking bone injury results obtained through SVM classification as antecedent input, and predicting the ligament injuries and hidden bone injuries. By means of the method, a surgeon can be assisted in automatically reading the ankle skeleton X-ray image, diagnosis and treatment time is saved, and the diagnosis and treatment effect of ankle injuries is improved.

Description

Technical field[0001]The present invention belongs to the field of image diagnosis, involving machine learning techniques, especially a machine-learning-based jet fracture and ligamental diagnosis method.Background technique[0002]The ankle joint needs to withstand the quality of 1.5 times when walking, and it is necessary to withstand the quality of 8 times when running, so the ankle joint in daily life is very susceptible to damage, accounting for 9% of the whole body fracture. In the clinical practice, the precision treatment of the ankle fracture is extremely difficult. One is the diversity of the type of the ankle fracture, the second is, in addition to damage the bone itself, the ligament will also be damaged, and most doctors can accurately identify if there is fracture, but experience Insufficient doctors are difficult to make accurate judgment on the position of the ankle injury and the type of injury. Due to irresponsible treatment of ankle surgery, the treatment of failure...

AI Technical Summary

Problems solved by technology

In 2019, Kitamura et al. (GeneKitamura, Chul Y.Chung, Barry E.Moore. Ankle Fracture Detection Utilizing a Convolutional Neural Network Ensemble Implemented with a Small Sample, De NovoTraining, and Multiview Incorporation[J]. Journal of Digital Imaging, 2019, 32 (4 ).) using convolutional neural networks (CNN), Pinto et al. (Daniel Pinto dos Santos, Sebastian Brodehl, Bettina Baeβler, Gordon Arnhold, Thomas Dratsch, Seung-Hun Chon, Peter Mildenberger, Florian Jungmann. Structured report data can be used to develop deep learning algorithms: a proof of concept in ankle radiographs [J]. Insights into Imaging, 2019, 10(1).) The identification of X-ray ankle fractures has been completed using Structural Reporting Data (SRD), but the current research on the automatic diagnosis of ankle fractures is still limited to The identification of the nature of the fracture, the type of classification in the field of machine learning cannot be used for the accurate classification of ankle fractures, and there is no research on the diagnosis of ligament injuries

Images